TRADE AND DEVELOPMENT REPORT, 2016

Transcription

1 UNITED NATIONS CONFERENCE ON TRADE AND DEVELOPMENT GENEVA TRADE AND DEVELOPMENT REPORT, 216 Structural transformation for inclusive and sustained growth Chapter III THE CATCH-UP CHALLENGE: INDUSTRIALIZATION AND STRUCTURAL CHANGE UNITED NATIONS New York and Geneva, 216

2

3 The Catch-up Challenge: Industrialization and Structural Change 57 Chapter III THE CATCH-UP CHALLENGE: INDUSTRIALIZATION AND STRUCTURAL CHANGE A. Introduction In recent years there has been a renewed interest in the role of industrialization in promoting sustained economic growth and development, reflected in Goal 9 of the 23 Agenda for Sustainable Development which calls for promoting inclusive and sustainable industrialization. Five important factors have contributed to this revival of interest. First, many developing countries have failed to deepen and diversify their existing industrial capacity in a more open global economy; indeed, several of them have experienced a premature decline in the share of manufacturing in their gross domestic product (GDP). Second, there is a perception that export-led growth strategies in developing countries face more constraints than in the past, in particular due to the slower growth of global demand, especially from industrialized countries. Third, many developing countries continue to remain vulnerable to external trade and financial shocks. Fourth, and related to the latter point, there has been an end to the enormous windfall gains from primary exports generated by the commodity price boom during the first decade of the 2s, which saw accompanying growth and investment spurts. And lastly, further deindustrialization in several developed countries is being observed with growing concern. 1 In the classic pattern of structural transformation, there is a decline in the relative share of the primary sector in GDP and a rise in the share of industry (which comprises, in addition to manufacturing, mining and quarrying, construction and utilities) (Kuznets, 1973). When a certain level of per capita income is reached, the share of industry in GDP stops growing while that of services rises. At the same time, industry s share of employment falls as productivity increases, even as the share of employment in services continues to rise. Historically, growth rates of industry have been closely related to those of GDP, and within industry, manufacturing has been critical. A broad and robust domestic manufacturing base has been the key to successful economic development, since it helps generate virtuous and cumulative linkages with other sectors of the economy, drives technological progress, and has the strongest potential for productivity gains. Thus, as manufacturing grows, primary production typically tends to become more efficient as a result of the greater use of capital and technology (including knowledge and technical skills) that not only contribute to productivity gains in manufacturing, but also to the development of the other subsectors of an economy. The services sector can emerge to supplement manufacturing activities from a certain level of per capita income onwards, and it may even grow to dominate the economy. However,

4 58 Trade and Development Report, 216 it is also possible that the expansion of services (in both output and employment terms) may occur even before a sufficiently diversified and dynamic industrial base is established, reflecting an interruption of the industrialization process. In developing countries, the drive towards manufacturing was originally based on the observation that these countries faced a structural disadvantage in global trade relations: as the prices of developingcountry exports (mainly primary commodities) tended to fall relative to those of developed-country exports (mainly manufactures) there was a decline in developing countries terms of trade, which tended to perpetuate the income gap between rich and poor countries (the Prebisch-Singer hypothesis, discussed also in chapter IV). Industrialization was expected to alter global trade patterns, resulting in changes in the international division of labour in a way that would be more favourable for developing countries. Diversification into manufacturing was expected to reduce developing countries dependence on the production and export of primary commodities and ease the balance-of-payments constraints on development by either replacing imports or by generating additional export earnings (Prebisch, 1964). Therefore, in the context of industrialization, this chapter gives particular emphasis to manufacturing, which, it is argued, is more likely to generate the linkages needed to sustain a virtuous circle of growth and structural transformation. Section B makes the case for the development of manufacturing as the means to structural transformation and income growth. Section C provides an assessment of changes in the economic structures of developing countries over the past few decades. Section D identifies different trajectories of structural change, and discusses why industrialization efforts, in terms of enlarging the share of manufacturing in total employment and value added, have been more successful in some developing countries than in others. Section E examines the potential contributions of the primary sector and services to the process of structural change. The final section draws a number of conclusions for policies in support of accelerated structural transformation. B. The case for developing manufacturing industries 1. The virtues of manufacturing The expansion of manufacturing activities can be considered as evolving through a process of cumulative causation (Myrdal, 1957; Kaldor, 1957, 1958) in which demand and supply factors interact: the expansion of manufacturing activities creates employment, incomes and demand, on the one hand, and accelerates increases in productivity on the other; this in turn boosts income and demand growth. Continuous upgrading of productive capacities in manufacturing, which is part of this process, can lead to productivity gains through entry into new areas of economic activity, the application of more advanced technologies, the production of more sophisticated goods, and/or the insertion into international value chains at rising levels of skill. 2 The immense appeal of manufacturing lies in its potential to generate productivity and income growth (Kaldor, 1966), and because such gains can spread across the economy through production, investment, knowledge and income linkages. Several linkages deserve mention here. To begin with, expanding production can help build backward linkages (to source inputs for production), and forward linkages insofar as the produced goods are used in other economic activities (Hirschman, 1958). Intersectoral linkages emerge as knowledge and efficiency gains spread beyond manufacturing to other sectors of the economy, including primary and service activities

5 The Catch-up Challenge: Industrialization and Structural Change 59 (Cornwall, 1977; Tregenna, 28; UNIDO, 29). Investment linkages are created when investments in productive capacity, new entrepreneurial ventures and the related extensions of manufacturing activities in one enterprise or subsector trigger additional investments in other firms or sectors, which otherwise would not occur because the profitability of a specific investment project in a certain area of manufacturing activity often depends on prior or simultaneous investments in a related activity (Rodrik, 24). Income linkages emerge from rising wage incomes generated from industrial expansion; these add to the virtuous cycle through consumption linkages. Income linkages also operate through supplementary government revenues (i.e. fiscal linkages ), which may therefore expand public expenditure (Hirschman, 1986). The creation of such income linkages can strengthen the self-reinforcing aspect of industrialization through increasing domestic demand and therefore GDP growth. Static economies of scale (i.e. lowering unit costs owing to increasing scale of production) tend to be substantial in manufacturing. In addition, there is scope for exploiting dynamic economies of scale when capital accumulation goes hand in hand with the use of increasingly sophisticated technologies, with knowledge acquisition through learning-by-doing and with the development of tacit skills and know-how. Spillovers of skill acquisition and technological learning across manufacturing firms, and from manufacturing to other sectors through both direct and indirect channels, in turn generate further productivity increases. A combination of these factors enables climbing the technology ladder through continuous upgrading of products, processes, organizational patterns and market possibilities (Schumpeter, 1961; Gerschenkron, 1962; Amsden, 21). Sustainable industrial catch-up and acceleration of structural transformation require a high rate of investment in productive capacity and technological Manufacturing activities create employment, incomes and demand, and accelerate productivity growth; this in turn boosts incomes and increases demand. Sustainable industrial catch-up and acceleration of structural transformation require a high rate of investment in productive capacity and technological capabilities. capabilities for several reasons. First, in order to benefit from static scale economies, firms need to increase their productive capacity. Second, efficiency gains that can result from improved allocation of factors of production and competition among manufacturing subsectors depend on the extent to which existing firms thrive and new firms emerge. This process is not possible without investment in new machinery and equipment. Third, productivity gains depend to a large extent on the introduction of new technologies that are embodied in machinery and equipment, which necessitates the periodic replacement of outdated machinery and equipment. These firm-level requirements add up to the need to increase investment rates and achieve some minimum level of per capita investment in the economy as a whole. Of particular importance is public investment in such crucial areas as transport and logistics, and telecommunications infrastructure, as well as the provision of power and water and other related utilities, as these indirectly boost the productivity of economic activities in all sectors and help accelerate the pace of structural transformation. The justification for the growth of manufacturing is not only economic, but also geopolitical and social. In today s globalized economy, a country that lacks a significant manufacturing sector may eventually face demand obstacles to growth and chronic balance-ofpayments constraints, making it vulnerable to decisions of external financial agents and to policy conditions set by official creditors. Moreover, industrialization allows the accumulation of technological capabilities that are important for any autonomous development strategy. In most developed countries, industrialization has also played a significant role in generating important social changes, including the expansion of a more formal organization of production and work. The developmental State an important actor in all successful industrialization experiences has played a catalytic role in promoting the rise of domestic entrepreneurs, and in fostering

6 6 Trade and Development Report, 216 the growth of urban, formal, increasingly skilled and potentially better organized working classes (see chapter VI). To the extent that these are associated with more cohesive and integrated societies, industrialization bolsters national consolidation and the stability of nation States, which in turn promote the development process (List, 1856). However, one caveat deserves to be pointed out: historically the expansion of manufacturing has tended to rely on patterns of production that damage the environment through pollution and lead to degradation and overexploitation of natural resources and excessive carbon emissions associated with climate change. Indeed, some observers have argued for a shift to services-based growth precisely in order to avoid the environmental problems that have emerged in some rapidly industrializing countries. But such problems are not intrinsic to the industrialization process: they depend crucially on the choice of technologies, as green technologies are now available for a wide range of manufacturing production processes (Pegels and Becker, 214). This also underlines the importance of facilitating the cheap and effective transfer of such technologies to developing countries (TDR 28). 2. Knowledge linkages and productivity growth The nature of technology and the knowledge acquisition for manufacturing change at different stages of industrial development. At the early stages, the skill levels required by the existing manufacturing subsectors may be relatively low, although on-the-job learning and experience can improve productivity. But from a certain stage of manufacturing development onwards, it is no longer advisable to rely solely on an abundant supply of low-skilled labour; adequately trained manpower and qualified personnel, including at various levels of management, become increasingly important. In order to be effective for sustained productivity and output growth, investment in productive capacity and technological upgrading therefore need to be combined with improvements and adaptation of workers skills, management knowhow and entrepreneurial competence. Knowledge acquisition refers to the accumulation of capabilities embodied in machinery and equipment as well as in people in the form of tacit know-how and skills (Lall, 1992, 23; Malerba, 22). Such knowledge contributes to productivity gains in two ways. First, access to already existing knowledge (information, tacit know-how and skills) helps enhance the efficiency and competitiveness of enterprises existing economic activities and processes. Second, the accumulation of new knowledge helps raise productivity, including through the introduction of new products, processes and organizational forms of doing business, which become more important as manufacturing output begins to expand. Such new knowledge supports further diversification of manufacturing activities, which in turn require a wider range of capabilities including through learning-by-doing and research and development (R&D) that promote innovation. Different kinds of manufacturing activities across various levels of technological intensity low, medium and high also have diverse implications for fostering further knowledge and skill acquisition. Generally, when learning takes place in design and engineering activities that feed a broader spectrum of sectors, industrial production leads to steep learning curves that promote greater intersectoral linkages and flows of knowledge. These can improve efficiency both in manufacturing and other related subsectors of the economy. In countries where advanced production technologies and new products are developed, the increase in the capital that embodies those technologies and the acquisition of skills on how best to use them advance in parallel. The situation is different in most developing countries, where technologies can be imported but the know-how and skills to optimize the use of such technologies have to be developed domestically. Moreover, imported technologies often have to be adapted to the specific requirements and possibilities of each country. Thus, developing countries that have a lower capacity to develop new technologies by themselves generally face the challenge of combining adaptation of available technologies with developing the know-how and skills for dealing with increasingly advanced technical equipment. While the use of acquired new knowledge in industry is an important source of upgrading, the dynamics are likely to be weaker in the case of participation in global value chains (GVCs) where technology-intensive inputs, product design and production processes are largely controlled by lead firms based outside the country or countries where the production takes place.

7 The Catch-up Challenge: Industrialization and Structural Change 61 Technological learning can occur at various levels, from school education and vocational training to learning-by-doing and R&D at the firm level, as well as in public or publicly supported institutions (Nübler, 214). As in the case of investment, where publicly provided infrastructure complements and is often a precondition for the viability of private investment, the public sector can make a crucial contribution to productivity growth by offering education, professional training and support to R&D. Equally important for learning are public information and coordination services that help private entrepreneurs assess the opportunities and risks of specific investment projects planned and undertaken by others, including in the area of public infrastructure (Rodrik, 24). In a dynamic process of upgrading in manufacturing, investment, technological advance and knowledge and skills acquisition are complementary: when there is an increase in one element, it also raises the marginal contributions of the others (Nelson and Winter, 1973; Dahlmann, 1979). Moreover, productivity growth is also cumulative over time, in that initial productivity increases in manufacturing activities generate further output and productivity increases. C. Trends in structural change since Long-term trends Over the past four and a half decades, the global economy as a whole has undergone significant changes in economic activities across sectors and regions. Developing countries increased their share of global industrial output (in current prices) from 15 per cent in 197 to 28 per cent in 22; it jumped dramatically thereafter to more than half by 214. Developing Asia accounted for two thirds of that increase. This shift in the distribution of industrial production to the advantage of developing countries resulted partly from an overall increase in their share of global output, and partly from the continuous decline of industry as a proportion of domestic value added in developed countries. In developed countries, the reduction of the share of industry in GDP was due almost entirely to manufacturing, the share of which fell from 26 per cent to 14 per cent of GDP. This decline was matched by the expansion of services, which since 29 have generated 75 per cent of these countries national income (chart 3.1). The transition economies and Latin America witnessed a similar trend of deindustrialization, with shares of services rising and those of industry and agriculture falling (in the latter case, from already relatively low levels). Asia presents a different picture. In this region, agriculture accounted for a significant proportion of GDP in 197: slightly more than 3 per cent in East and South-East Asia, 4 per cent in South Asia and 2 per cent in West Asia. By 214, the share of this sector had declined by between 25 and 15 percentage points in all these subregions. The main counterpart was the increase in the share of services, while the weight of industry remained roughly constant, or increased slightly as in South-East Asia. In West Asia, the share of industry has remained high and has even increased in recent years, largely driven by mining and the effect of higher global prices of crude oil and natural gas. Finally, in Africa as a whole there has been little change in the production structure, with the share of agriculture declining only slightly to the moderate benefit of industry, while the share of the services sector has remained stable at around 5 per cent of GDP. In terms of employment, most developing regions have experienced a sizeable increase in the share of industrial employment since 197 (table 3.1), although for many countries this has been due to construction rather than to manufacturing. The

8 62 Trade and Development Report, 216 Chart 3.1 SHARE OF ECONOMIC SECTORS IN TOTAL VALUE ADDED, BY COUNTRY GROUP, (Per cent at current dollars) Developed economies Africa East Asia South Asia Transition economies Latin America and the Caribbean South-East Asia West Asia Agriculture, hunting, forestry, fishing Industry Services Source: UNCTAD secretariat calculations, based on UNCTADstat.

9 The Catch-up Challenge: Industrialization and Structural Change 63 Table 3.1 SHARE OF INDUSTRY IN TOTAL VALUE ADDED AND EMPLOYMENT, SELECTED GROUPS AND ECONOMIES, (Per cent) Share of industry value added At current dollars At constant dollars Share of industry in total employment Developed economies North Africa Sub-Saharan Africa South Africa Latin America and the Caribbean Argentina Brazil Chile Mexico East Asia China Republic of Korea South-East Asia Indonesia Malaysia Philippines Thailand South Asia India West Asia Transition economies Source: UNCTAD secretariat calculations, based on UNSD; ILO, Key Indicators of the Labour Market: KILM 4; The Conference Board, Total Economy Database; World Bank, World Development Indicators database; Groningen Growth and Development Centre, GGDC- Sector Database. Note: Calculations at constant prices are based on value added at constant 25 dollars. Regional values correspond to unweighted averages. Industry comprises sectors C-F of ISIC Rev. 3. The samples of economies by country group are as follows: Developed economies: Australia, Austria, Belgium, Bulgaria, Canada, Croatia, Cyprus, the Czech Republic, Denmark, Estonia, Finland, France, Germany, Greece, Hungary, Iceland, Ireland, Israel, Italy, Japan, Latvia, Lithuania, Luxembourg, the Netherlands, New Zealand, Norway, Poland, Portugal, Romania, Slovakia, Slovenia, Spain, Sweden, Switzerland, the United Kingdom and the United States of America. North Africa: Egypt, Morocco and Tunisia. Sub-Saharan Africa: Botswana, Ethiopia, Ghana, Kenya, Malawi, Mauritius, Nigeria, Senegal, South Africa, the United Republic of Tanzania and Zambia. Latin America and the Caribbean: Argentina, Barbados, the Bolivarian Republic of Venezuela, Brazil, Chile, Colombia, Costa Rica, Ecuador, El Salvador, Guatemala, Mexico, Peru, the Plurinational State of Bolivia, Trinidad and Tobago and Uruguay. East Asia: China, Macao (China), the Republic of Korea and Taiwan Province of China. South-East Asia: Indonesia, Malaysia, the Philippines, Singapore and Thailand. South Asia: Bangladesh, India, Pakistan and Sri Lanka. West Asia: Bahrain, Jordan, Saudi Arabia, the Syrian Arab Republic and Turkey. Transition economies: Albania, Armenia, Azerbaijan, Belarus, Georgia, Kazakhstan, Kyrgyzstan, the former Yugoslav Republic of Macedonia, Republic of Moldova, the Russian Federation and Ukraine.

10 64 Trade and Development Report, 216 Chart 3.2 SHARE OF MANUFACTURING IN TOTAL VALUE ADDED, BY COUNTRY GROUP, (Per cent) 4 A. At current dollars 4 B. At constant 25 dollars Developed economies Transition economies Latin America and the Caribbean Africa East Asia South-East Asia South Asia West Asia Source: UNCTAD secretariat calculations, based on UN DESA, Statistics Division, National Accounts Main Aggregates database. exception is Latin America and the Caribbean, where the share of industrial employment has remained virtually stable. Other than in sub-saharan Africa, industry in the other developing regions/subregions accounts for between 2 and 3 per cent of total employment. Within the industrial sector, with the exception of East Asia, there was a general reduction in the share of manufacturing in value added (at current prices), although this started at different points in time (chart 3.2A). In 197, only the transition economies and East Asia exhibited a higher share of manufacturing in GDP than developed countries. But the situation has reversed since then, and after 2 only in West Asia and Africa were the shares of manufacturing in GDP clearly smaller than those of the developed economies. From this point of view, it would seem that most developing economies have narrowed the industrialization gap, with the share of manufacturing in total value added closer to or even higher than that of developed economies, although this convergence has been occurring within an overall declining trend. This structural change had already started for the developed countries in the 196s and 197s, with a secular decline in the share of manufacturing employment (table 3.2). Since then, deindustrialization has spread to developing countries. Contributory factors to this general trend include financialization in the global economy, which generated macroeconomic instability and increasing inequality in income distribution. This has contributed to the slowdown of aggregate demand in the context of stagnating wage incomes and low quality and informal employment, which are associated with weaker productivity performance, underconsumption and lower levels of investment (see chapter II). This downward tendency also reflects a broader global trend of falling prices of manufactured goods relative to the general price level, resulting, in particular, from faster productivity growth. Hence, measured at constant prices, the decline in the share of manufacturing in GDP in several regions has been much less steep over time, whereas that share has continued to grow in some Asian regions (chart 3.2B). The even sharper increase in the share of manufacturing in East Asia in constant prices (driven mainly by China) suggests an additional reason for this trend:

11 The Catch-up Challenge: Industrialization and Structural Change 65 Table 3.2 SHARE OF MANUFACTURING IN TOTAL VALUE ADDED AND EMPLOYMENT, SELECTED GROUPS AND ECONOMIES, (Per cent) Share of manufacturing in total value added Share of manufacturing At current dollars At constant dollars in total employment Developed economies North Africa Sub-Saharan Africa South Africa Latin America and the Caribbean Argentina Brazil Chile Mexico East Asia China Republic of Korea South-East Asia Indonesia Malaysia Philippines Thailand India Source: UNCTAD secretariat calculations, based on UNSD; and Groningen Growth and Development Centre, GGDC- Sector Database. Note: Calculations at constant prices are based on value added at constant 25 dollars. Regional values correspond to unweighted averages. Manufacturing corresponds to sector D of ISIC Rev. 3. The samples of economies by country group are as follows: Developed countries: Denmark, France, Germany, Italy, Japan, the Netherlands, Spain, Sweden, the United Kingdom and the United States of America. North Africa: Egypt and Morocco. Sub-Saharan Africa: Botswana, Ethiopia, Ghana, Kenya, Malawi, Mauritius, Nigeria, Senegal, South Africa, the United Republic of Tanzania and Zambia. Latin America and the Caribbean: Argentina, the Bolivarian Republic of Venezuela, Brazil, Chile, Colombia, Costa Rica, Mexico, Peru and the Plurinational State of Bolivia. East Asia: China, the Republic of Korea and Taiwan Province of China. South-East Asia: Indonesia, Malaysia, the Philippines, Singapore and Thailand.

12 66 Trade and Development Report, 216 the dramatic increase in volumes of manufactured goods in this subregion must have played a role in the overall decline in the relative prices of manufactures. In developed countries, the shares of manufacturing in both total employment and GDP first increased with the expansion of real income, then peaked at some point, after which they started to decline, with a concomitant increase in the share of services, following an inverse U-shaped curve. The phase of deindustrialization is a normal consequence of the development process, owing to changes in the composition of demand and greater productivity gains in manufactures than in most other economic sectors (TDR 23). The dynamics of demand, which at earlier stages of development encourage industrialization by rapidly expanding the demand for manufactures, tend to favour the expansion of services as income levels keep growing. 3 The relatively high productivity in manufacturing leads to a reduction in the share of that sector in total employment, followed by a reduction in total value added, particularly in nominal terms, as higher productivity gains in manufacturing tend to reduce the relative prices of manufactures. This combination of demand and technological factors explains why, in the developed countries, the share of manufacturing in employment peaked before its share in nominal value added, which in turn preceded its peak in real value added (Rodrik, 215). With the exception of East Asia, there has been a general reduction in the share of manufacturing in total value added. However, as UNCTAD has also noted (TDRs 1995, 23), deindustrialization in developed economies, particularly in some European economies, has not been completely smooth and spontaneous to the extent that it has been associated with institutional and financial transformation and regressive income distribution. These factors slowed down the growth of aggregate demand and constrained the capacity of services to productively absorb labour released from industry, leading to higher and persistent underemployment or unemployment rates (Palma, 25). 4 Even so, such a process of deindustrialization in developed economies has generally occurred when the prior process of industrialization had already raised overall productivity in the economy, disseminated technological capacities and consolidated a domestic market. The experience of developing and transition countries has been much more varied. The transition economies experienced the most dramatic reduction in the share of manufacturing in GDP in the second half of the 198s and in the 199s; indeed, manufacturing was particularly affected by the economic crisis. The subsequent recovery of GDP growth following the 1998 crisis in the Russian Federation benefited all sectors, and led to the stabilization of the share of manufacturing in GDP at constant prices (chart 3.2B). By the 197s, countries in the Latin America and Caribbean region had developed substantial industrial capacity, including in manufacturing in some countries such as Argentina and Brazil (table 3.2). Thereafter, there were steep declines in the shares of manufacturing in value added. The abandonment of long-standing industrialization strategies, beginning in the Southern Cone in the late 197s, followed by aggressive structural adjustment were clearly factors contributing to such an outcome. The declining trends were generalized throughout the region following the debt crisis in the 198s and the policy conditionalities imposed by the International Monetary Fund (IMF) and the World Bank. In some countries, deindustrialization trends were temporarily masked by large real devaluations in the late 198s and late 199s, as higher relative prices of manufactures (among other tradable goods) concealed falling production in real terms. Conversely, after 23, real exchange rate appreciation accentuated the reduction of the share of manufacturing in GDP at current prices, even though growth in manufacturing recovered and the decline in manufacturing value added was less marked (in constant price terms). The steep variations in the shares of manufacturing in total value added in West Asia can also be partially explained by shifts in relative prices (chart 3.2A). Thus, the declining share of manufacturing in the 197s was not because of low real growth rates of the sector (which averaged around 7 per cent per annum), but rather because of the huge increase of the mining sector s share in nominal terms as a result of rising oil prices, which reduced the shares of all the other sectors. The subsequent increase in the share of manufacturing until the late 199s was

13 The Catch-up Challenge: Industrialization and Structural Change 67 partly due to greater dynamism in this sector, especially in Turkey, and partly reflected a reversal in oil prices. Similarly, the region experienced significant deindustrialization during the 2s (measured as a share of GDP at current prices). However, during this period the growth of manufacturing accelerated to around 6 per cent per annum and its share in GDP increased slightly (at constant prices, chart 3.2B). Africa remains largely dependent on the primary sector, with a low share of manufacturing in GDP, fluctuating between 12 per cent and 15 per cent (at current prices) until the 2s. However, at constant prices, the fluctuations have been less pronounced. Since 28, the share of manufacturing has stagnated at around per cent of GDP, at both constant and current prices (chart 3.2). It is worth noting that this has occurred in the context of a significant acceleration of manufacturing production in the region. In sub- Saharan Africa (excluding South Africa), the growth rates of manufacturing jumped from an annual average of.2 between 199 and 2 to 5 per cent between 21 and 28, and to 7.6 per cent between 29 and 214. This does not necessarily indicate that a sustained process of industrialization is under way, since the starting point was low. For this to happen, manufacturing growth should be supported by a significant expansion of investment, and should last long enough to trigger the dynamics of structural transformation. Most Asian countries present a rather different picture. The shares of manufacturing in GDP continued to grow at current prices until the mid-2s in South-East Asia and until 2 in East Asia, and are now the highest in the world. At constant prices, these shares have grown even faster in East Asia or decreased slightly in South-East Asia, reflecting the change in relative prices of manufactures, noted earlier, and pointing to the significance of volumes of production emanating from East Asia. Similarly, the share of manufacturing in South Asia shows a marginal increase and then a decline at current prices, but an increase at constant prices. Nevertheless, that share remains relatively low by both measures, at around 17 per cent of GDP. These examples illustrate that a proper evaluation of industrialization or deindustrialization processes Industrialization processes depend on the strength of the investment drive and the generation of production, income and knowledge linkages. must be based on an understanding of their broader economic context. It is evident that a falling ratio of manufacturing in value added may reflect the absolute regression of that sector, with an associated loss of production capabilities, knowledge and expertise and the weakening of production linkages. Or it could simply result from the fact that, even if it is growing, other sectors are growing faster. The nature and implications of these different processes cannot be assessed without taking into account the existence or absence of a strong investment drive to support sustained economic growth, and the generation of productive, income and knowledge linkages, as discussed in the remainder of this chapter. 2. Impact of structural change and investment on aggregate productivity (a) Productivity growth and structural change As noted above, changes in the sectoral composition of output and employment have to be seen in relation to patterns of investment, growth and productivity. With regard to the crucial macroeconomic elements of structural transformation since 197, namely the growth of GDP, industry value added, employment, labour productivity and investment, there were substantial differences across developing regions, but also a marked contrast between the preand post-198 periods for most groups of developing economies (tables 3.3 and 3.4). Overall, during the 197s the majority of developing economies experienced some structural change, supported by industry output and employment growth and also by increased labour productivity. 5 Since the 197s, except for East Asia and South Asia (and sub-saharan Africa in the post-2 period), no developing region/subregion was able to maintain annual GDP growth rates at similar levels to those experienced in the 197s. GDP growth in Latin America, West Asia and North Africa fell sharply in the 198s with concomitant slumps in industrial output growth, demonstrating thereby the close connection also between value-added growth in industry and the overall growth rate of an economy. 6

14 68 Trade and Development Report, 216 Table 3.3 AVERAGE ANNUAL GROWTH RATES OF TOTAL VALUE ADDED, VALUE ADDED IN INDUSTRY AND TOTAL EMPLOYMENT, SELECTED GROUPS AND ECONOMIES, (Per cent) Total value added growth Industry value added growth Employment growth Developed economies North Africa Sub-Saharan Africa South Africa Latin America and the Caribbean Argentina Brazil Chile Mexico East Asia China Republic of Korea South-East Asia Indonesia Malaysia Philippines Thailand South Asia India West Asia Saudi Arabia Turkey Transition economies Source: See table 3.1. Note: Calculations are at constant 25 dollars or number of employees. For the country samples in the groups, see table 3.1.

15 The Catch-up Challenge: Industrialization and Structural Change 69 Table 3.4 AVERAGE ANNUAL GROWTH RATES OF INVESTMENT, TOTAL LABOUR PRODUCTIVITY AND LABOUR PRODUCTIVITY IN INDUSTRY, SELECTED GROUPS AND ECONOMIES, (Per cent) Investment growth Labour productivity growth Industry labour productivity growth Developed economies North Africa Sub-Saharan Africa South Africa Latin America and the Caribbean Argentina Brazil Chile Mexico East Asia China Republic of Korea South-East Asia Indonesia Malaysia Philippines Thailand South Asia India West Asia Saudi Arabia Turkey Transition economies Source: See table 3.1. Note: Calculations are at constant 25 dollars or number of employees. For the country samples in the groups, see table 3.1.

16 7 Trade and Development Report, 216 Uneven patterns of output growth, employment generation and productivity dynamics emerged in developing countries from the 198s onward. These provide another important insight: that rates of growth of GDP, investment, industry, employment and productivity have all tended to move together in cases of successful structural change. There were similar growth rates of industrial employment for several countries, but Asian countries already showed considerably higher rates of growth of output and employment in industry (table 3.3), supported by rapid productivity growth in industry (table 3.4). Within any economy, productivity levels can vary considerably, depending on the economic activity, the size of the firm and the degrees of formality and informality of employment. Labour productivity tends to be particularly high in the mining sector (which includes hydrocarbons), as the bulk of mining production is undertaken by large, capitalintensive firms. Typically, the productivity level of manufacturing tends to be well above the national average, although this varies, as microenterprises and informal jobs displaying relatively lower productivity coexist with large firms that use high-technology and skilled labour and therefore have higher productivity. The lowest output per worker is generally in agriculture, especially in Asian and African countries where most producers are small peasants who use less mechanized technologies. Finally, the productivity of services depends on the type of activity. In general, finance, insurance, real estate, business services, and transport, storage and communications have relatively high levels of productivity (though initially this may simply reflect price movements), while community, social and personal services and government services tend to have much lower output per worker. Productivity in trade, restaurants and hotels is quite varied, but tends to be rather low in most developing countries, where a large segment of informal commerce exists. Such variation suggests that aggregate productivity can be enhanced by reallocating employment from lower productivity to higher productivity activities, both within and between sectors. Clearly, the potential for this is greater in countries and regions where much of the labour force is employed Growth rates of GDP, investment, industry, employment and productivity have all tended to move together in cases of successful structural change. in low-productivity activities, as is generally the case in Africa and Asia today, and was the case in Latin America around 195. This potential was well exploited in Latin America until 198, when the decline in the share of agriculture in total employment (from 55 per cent in 195 to 32 per cent in 198) was matched by an increase in shares of all the other sectors. Since 198, however, the share of agriculture in total employment has been further declining, with employment redistributed mostly to low-productivity services. In a sample of Latin American countries covering most of regional output and population, aggregate productivity increased steadily until 198, but has stagnated or declined in almost all sectors since then (chart 3.3A), reflecting, inter alia, a weakening of investment. African countries also managed to increase aggregate productivity until 198, thanks to a combination of productivity growth in industry and modern services, and to some though limited reallocation of employment from agriculture to (mostly) other services. Even though productivity levels in these services were relatively low, they were nevertheless much higher than in agriculture (chart 3.3B). These factors weakened or disappeared between 198 and 2, as agriculture stopped losing its share of employment, and productivity in most modern sectors (with the exception of transport and communications) slowed down. There was a recovery in the 2s, with a moderate decline in the share of agriculture in total employment along with some improvements in productivity, mostly in agriculture and low-productivity services. Since these sectors still employ 82 per cent of the population in these countries, any improvement in their productivity levels is of macroeconomic significance. By contrast, output per worker in mining is around 2 times the average, but it employs less than 1 per cent of the labour force in these countries. Over the past half century, Asian countries experienced the greatest structural change as well as a stronger increase in productivity levels, although these started from very low levels (chart 3.3C). Whereas in the early 196s, agriculture accounted for 77 per cent of the region s total employment, by 2 this had fallen to 42 per cent, largely due to

17 The Catch-up Challenge: Industrialization and Structural Change 71 Chart 3.3 EMPLOYMENT, VALUE ADDED AND PRODUCTIVITY BY ECONOMIC SECTOR IN SELECTED COUNTRY GROUPS, VARIOUS YEARS (Per cent and constant PPP dollars per employee) Employment share by sector (Per cent of total employment) Other services FIRE, TSC Other industries Manufacturing Agriculture Value added share by sector (Per cent of total value added) A. Latin America B. Africa C. Asia Other services FIRE, TSC Other industries Manufacturing Agriculture Productivity by sector (Thousands of 25 PPP dollars) Other services FIRE, TSC Other industries Manufacturing Agriculture Total Source: UNCTAD secretariat calculations, based on Groningen Growth and Development Centre, GGDC- Sector Database. Note: FIRE = finance, insurance, real estate and business services, TSC = transport, storage and communications both categories represent higher productivity service groups. Other services comprise community, social and personal services and government services, as well as trade, restaurants and hotels which are relatively lower productivity groups. Other industries comprise mining and quarrying, construction and utilities. Calculations are based on weighted regional averages for the sample of economies, as listed below. Africa: Botswana, Egypt, Ethiopia, Ghana, Malawi, Morocco, Nigeria, South Africa, the United Republic of Tanzania and Zambia; Asia: China, India, Indonesia, the Republic of Korea, Taiwan Province of China and Thailand; Latin America: Argentina, the Bolivarian Republic of Venezuela, Brazil, Chile, Colombia, Costa Rica, Mexico and the Plurinational State of Bolivia.